AU2002223560A1

AU2002223560A1 - Live influenza vaccine and method of manufacture

Info

Publication number: AU2002223560A1
Application number: AU2002223560A
Authority: AU
Inventors: Andre Egorov; Boris Ferko; Dietmar Katinger; Hermann Katinger; Julia Romanova
Original assignee: Polymun Scientific Immunbiologische Forschung GmbH
Current assignee: Polymun Scientific Immunbiologische Forschung GmbH
Priority date: 2000-09-25
Filing date: 2001-09-25
Publication date: 2002-06-20
Anticipated expiration: 2021-09-25

Description

LIVE VACCINE AND METHOD OF MANUFACTURE

TECHNICAL FIELD

The present invention is in the field of virology and vaccine development and relates to an improved method of manufacture of a viral vaccine, particularly of a whole-virus vaccine, preferably of an attenuated live vaccine and to vaccines obtainable by the method.

BACKGROUND OF THE INVENTION

The influenza hemagglutinin (HA) antigen is the major target for the protective immune responses of a host to the virus.

A common practice of recovering new viral isolates involves recovery from a nasal or throat swab or from a similar source, followed by cultivation of the isolates in embryonated chicken eggs. The virus adapts to its egg host and large scale production of the virus can be carried out in eggs. Such conventional methodology involving embryonated chicken eggs to produce Influenza vaccine is, however, extremely cumbersome, involving the handling of many thousands of eggs per week as well as extensive purification of the virus suspension derived from the allantoic fluid to ensure freedom from egg protein.

Another disadvantage in the use of chicken embryos for virus production lies in the fact that this substrate strongly favors the selection of virus variants that differ in their antigenic specificity from the wildtype virus and not rarely results in viruses that may not be suitable for vaccine production due to their altered phenotypes including, for instance, considerable reduction in immunogenicity.

Many attempts have therefore been undertaken in the art to utilize standard tissue culture technology with established mammalian cell lines, such as MDCK (Madin-Darby Canine Kidney) or Vero (African Green Monkey Kidney) cells, for virus production, particularly influenza virus production.

One of the difficulties in growing influenza strains in tissue cell culture arises from the necessity for proteolytic cleavage of the influenza hemagglutinin in the host cell. Cleavage of the virus HA precursor into the HA1 and HA2 subfragments, although not necessary for the assembly of the viral elements to form a complete virion, is required, however, to render the virion infective, i.e. to enable it to infect a new cell.

It has been reported (e.g. Lazarowitz et al., " Enhancement of the Infectivity of Influenza and B Viruses by Proteolytic Cleavage of the Hemagglutinin

Polypeptide", Virology, 68:440-454, 1 975) that the limited replication of several influenza A strains in standard cell cultures could be overcome by the addition of proteases like trypsin to the tissue culture medium. Yet, there remained difficulties in some cases, for instance when using Vero cells.

Kaverian and Webster (J Virol 69/4:2700-2703, 1 995) report that in Vero cell cultures, and less pronounced in MDCK, swine kidney, or rhesus monkey kidney cell cultures, the trypsin activity in the medium rapidly decreased from the onset of incubation resulting in the failure of virus accumulation in the medium due to the lack of production of a sufficient number of infective virions. They concluded that a trypsin inhibiting factor was released from the Vero cells. They further showed that by repeated addition of trypsin reproduction of virus could be resumed and maintained for a number of reproduction cycles resulting in a much better virus yield.

Another way for efficient vaccine production was reported in US 5,753,489 wherein serum-free medium was used for virus propagation in a number of different mammalian cells including MDCK and Vero cells. The method disclosed therein comprises growing vertebrate cells in serum-free medium, infecting the cell culture with a virus, incubating the cell culture infected with the virus, removing a portion of the virus-containing medium and contacting this portion with a protease, thereafter adding to that portion a protease inhibitor and returning that portion to the cell culture. It is preferred therein to provide the steps of growing, infecting and incubating in a first vessel and the steps of trypsin-contacting and inhibitor-adding are performed in a second vessel connected with the first vessel in a loop so that the steps o can be performed in a closed cycle. This system allows to use trypsin or other proteolytic enzymes at much higher concentrations than those normally tolerated by cells in culture.

EP 0870508 reports a method to produce a viral antigen vaccine comprising infecting an animal cell line, optionally a Vero cell line, with virus, propagating virus in the cell culture, adding a nuclease enzyme to the cell culture shortly before the end of virus propagation to digest nucleic acid material released from the lysing host cells into the medium, harvesting the virus and obtaining viral antigens thereof by extraction in order to make the viral antigen vaccine. The patent is silent with regard to the kind of nutrient medium used for virus propagation and also with regard to the addition of a protease, usually required for the final processing of influenza virus hemagglutinin to get infectious virus. The method further requires various purification steps for providing a ready-for- use vaccine preparation.

It is known, however, that the nature the host substrate as well as the composition of the nutrient medium used for virus propagation may significantly affect immunogenicity and antigenicity of the virus progeny obtained therewith. Particularly, serum-containing media may not only decrease antigenicity of viral progeny but additionally may decrease protease activity in the medium, hence inhibit virus maturation, and subsequently require expensive steps of purification.

SUMMARY OF THE INVENTION

The present invention overcomes the drawbacks of the prior art. It relates to a simple and efficient process for isolating viruses from various sources and for producing viral progeny for use as vaccines, particularly live attenuated influenza vaccines, in under conditions where alterations in the surface antigens of the virus due to adaptive selection are minimized or entirely prevented.

It is also an object of the present invention to provide for a method for the production of viruses, particularly influenza viruses, that yields viral progeny that selectively agglutinates human erythrocytes but not chicken erythrocytes, and that preferably has antigenic properties identical with those of the initially inoculated virus strain, e.g. a primary clinical wildtype isolate.

In a preferred embodiment, the nucleic acid sequence of the HA gene and optionally of the NA gene of the propagated virus is identical with the one of the initially inoculated strain (e.g. an epidemic strain, primary clinical isolate of an infected patient). It is another object of the invention to provide a method for efficient production of a whole-virus vaccine, particularly a live attenuated vaccine, in a single step procedure that does not require any chromatographic or other purification steps of the virus suspension harvested from the cell culture supernatant by centrifugation, particularly no protein separation or purification steps.

It is yet another object of the invention to provide attenuated, cold adapted and temperature sensitive influenza A and B strains and vaccines made thereof.

BIREF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a graphic illustration of the time course of trypsin inactivation in the supernatant of a Vero cell culture.

Fig. 2 is a graphic illustration of the time course of trypsin inactivation in the supernatant of a MDCK cell culture.

DETAILED DESCRIPTION OF THE INVENTION

Comparative experiments using embryonated eggs, MDCK and Vero cells clearly proved that the initially inoculated virus is likely to undergoe antigenic alteration during growth on any one of these substrates

Our experiments confirmed that the alterations are least or even absent for influenza virus strains grown on Vero cells in serum-free medium. Moreover, it turned out that influenza A viruses,, at least strains of the H3N2 subtype, when multiplied on Vero cells in serum-free and protein-free medium exhibit a selectivity for agglutination of human erythrocytes but not for chicken erythrocytes. Also, they did not grow on eggs. This was a first indication that these Vero-grown viruses might be more identical with the wildtype virus of the corresponding clinical isolate than the ones grown on MDCK cells or eggs.

Indeed, comparison of the HA and NA gene sequences of wildtype isolates obtained from nasal swabs with the ones of the same viruses after growth on Vero and MDCK cells, respectively, revealed alterations in the HA or NA of MDCK-grown viruses relative to the HA or NA of the swab isolates or of the Vero-grown viruses or of both the swab isolates and the Vero-grown viruses. Moreover, experimental data obtained from immunizations of ferrets with Vero- and MDCK-grown wildtype viruses indicate a far stronger virulence of the Vero- grown viruses compared to the MDCK-grown viruses. Also, the immunogenicity of the Vero-grown viruses tested in an animal trial on macaques was demonstrated to be significantly superior to the one of the viruses grown on MDCK cells or eggs.

These findings together provide strong evidence for the hypothesis that the process for the multiplication and propagation of viruses according to the present invention as hereinafter described in more detail yields viruses that are either unaltered compared to the initially inoculated (e.g. wildtype) virus or are modified to only a minor extent.

It is not only the avoidance of antigenic alterations that makes the present process of virus multiplication so unique, but it is also its striking simplicity which makes it extremely suitable for large scale industrial vaccine production.

Further experiments have shown that the source of trypsin (or trypsinogen) may be one additional factor that influences the overall yield of infective virions. Indeed, while the methods known in the art (e.g, Kaverin and Webster, J Virol 69/4:2700-2703, 1 995; or US 5,753,489) use either repeated addition of trypsin (Kaverin and Webster) or high trypsin concentrations (US 5,753,489), the process according to the present invention applies only half or less of the trypsin concentrations reported in the prior art. Moreover, a single addition of as little as 0.5 - 10 μg, preferably 2 - 5 μg trypsin per ml to the cell culture medium prior to or at the beginning of incubation of the infected host cells is sufficient to reach optimal infective virus titers. Inactivation experiments revealed that porcine or human recombinant trypsins are far less susceptible to inactivation by Vero or MDCK cells than bovine trypsin. Since bovine trypsin is most commonly used in the art it is rather likely that prior art literature unless explicitly mentioning another trypsin source, implicitly refers to bovine trypsin only. This would also help to explain the modes and concentrations of trypsin application recited, for instance, in Kaverin et al. and in US 5,753,489.

Using porcine or human rec trypsin or trypsinogen for initially supplementing the serum-free medium for Vero cell cultures according to the present invention therefore allows to use extremely low trypsin or trypsinogen concentrations and thus prevents the need of labor-intensive and costly purification steps after harvesting of the virus-containing supernatant.

Another step that contributes to make the present process simple and therefore attractive to vaccine manufacturers is the addition of a single dose of highly active endonuclease to the cell culture medium prior to or at the beginning of incubation of the infected Vero cells for virus propagation. This endonuclease, preferably Benzonase™, is added once to the medium at a very low initial concentration of 2 - 30, preferably 5 - 1 5, Units per ml of medium and effectively clears the cell culture medium from free DNA and RNA originating mainly from the lysing or lysed host cells. The residual Benzonase enzyme concentration in the ready-for-use vaccine preparations obtained from the centrifuged supernatant remains at 5 ng or less per dose.

Benzonase™ is a trademark of Nycomed Pharma A/S Denmark and relates to an extracellular unspecific endonuclease obtained from Serratia marcescens. Benzonase is a genetically engineered endonuclease which degrades both DNA and RNA strands in many forms to small oligonucleotides. It promotes quick reduction of the viscosity of cell lysates, which facilitates ultracentrifugation. It reduces proteolysis and increases the yield in targeted protein and offers complete elimination of nucleic acids from, e.g. recombinant, proteins. It has an exceptionally high activity of 400,000 U/mg.

A third and important advantage of the present process is the factor time hence process costs. Due to the use of serum-free medium that does not contain proteins of animal origin and preferably no antibiotics, expensive and time- consuming purification procedures can be reduced to a minimum or even totally avoided. Also, because the addition of exogenous enzymes such as the protease (e.g. trypsin or trypsinogen) and nuclease (e.g. Benzonase) occurs once at the beginning of the virus propagation phase this saves plenty of time that the state-of-the-art methods require for post-incubation treatment of the virus-containing culture supernatant (e.g., HA activation, RNA/DNA digestion, protein purification, etc.).

Surprisingly, it turned out that the early addition of either or both of protease (e.g. trypsin or trypsinogen) and nuclease (e.g.Benzonase) to the virus-infected Vero-cell culture had no negative implications on the virus yield, which is probably due to the very low enzyme concentrations applicable in the process of the present invention.

The present process of virus propagation is useful for the multiplication of various kinds of viruses, particularly influenza A viruses of the H3N2 subtype, but is also suitable for the isolation and reproduction of any epidemic or laboratory influenza virus strain, regardless of the kind of virus inoculum (e.g., blood serum sample, nasal wash, nasal swab, pharyngeal swab, saliva, etc.). Using the principles of this process, a number of influenza A and B vaccines has been produced which are part of the present invention and which are characterized in more detail in the subsequent Examples.

Also, protective efficacy as well as vaccine safety have been confirmed for the vaccines made according to the present invention, as will be demonstrated in the Examples.

The term "protein-free" or "free of non-serum proteins" as used herein in connection with the method of virus multiplication or propagation according to the present invention shall mean free of any functionally active protein. It shall not exclude, however, non-functional peptides as may originate from protein hydrolysates such as yeast extract or soya extract. Unless stated otherwise, the term " protein-free" shall neither exclude the presence of a protease and a nuclease enzyme at the concentrations disclosed and claimed herein.

In a preferred embodiment, the present invention relates to a simple, reliable and highly economic method for the manufacture of a whole-virus vaccine, preferably of an attenuated live vaccine, comprising the steps of: a) infecting African Green Monkey Kidney (Vero) cells with a desired virus, wherein the Vero cells have been grown in and separated from a serum-free medium that is also free of non-serum proteins; b) combining the infected cells with a suitable serum-free cell culture medium that is also free of non-serum proteins except for a protease and a nuclease; and c) incubating the cells in the presence of said protease and said nuclease to allow for production of infectious virus and, simultaneously, for digestion of nucleic acid material released to the cell culture medium; d) harvesting infectious virus by collecting virus-containing supernatant obtained from centrifugation of the cell culture; and e) preparing a vaccine thereof comprising subjecting the virus-containing supernatant to at least one processing step selected from the group consisting of filtering, concentrating, freezing, freeze-drying, and stabilizing by addition of a stabilizing agent.

It is preferred that the virus used for propagation has never had any contact to a host substrate other than a Vero cell line. This will ensure best results with regard to immunogenic and antigenic identity of the initial virus (e.g. nasal swab isolate) and the viral progeny obtained after propagation.

It is also preferred that the virus used for propagation, particularly for the manufacture of a whole-virus vaccine, preferably an influenza attenuated live vaccine, is an influenza virus selected from the group consisting of strains A/Sing/1 /57ca, A/Sing/1 /57ca/ΔNS 87, A/Sing/1 /57ca/ΔNSPR8, A/Sing/1 /57ca/NS1 24PR8, B/Vienna/1 /99ca, B/Vienna/99/ca37 and any attenuated variants and reassortants derived from any one of these strains. The genetic characteristics of the preferred virus strains, e.g. master strains, are disclosed in full detail in the subsequent Examples.

In another embodiment, the present invention refers to a whole-virus vaccine itself, preferably to an attenuated live vaccine, which in its ready-for-use form comprises essentially unmodified, optionally filtered and/or concentrated, virus- containing supernatant of a serum-free and protein-free Vero cell culture used for production of said virus. It is particularly preferred that the vaccine is produced according to the method of the present invention as disclosed and claimed herein.

This "one-step" vaccine, which does not require further processing, e.g., purification steps other than centrifugation and/or conventional filtration (i.e. not gel filtration), is compliant with the requirements for FDA approval.

The term "essentially unmodified" as used herein with regard to virus-containing supernatant in vaccine preparations according to the present invention shall refer to the composition of the supernatant as is at the time of harvesting the propagated virus, i.e. to the composition of the soluble components and ingredients present in the liquid phase of the supernatant. Minor alterations of the composition of ingredients as may occur due to steps of, for example, filtration, sterile filtration, centrifugation, concentration, drying, or freeze-drying of the virus-containing supernatant, shall be regarded as falling within the scope of "essentially unmodified" . Also, the term shall not exclude the presence of preserving and/or stabilizing agents usually applied in the art to vaccine preparations.

The whole-virus vaccines of the present invention may be used for the prophylactic or therapeutic treatment of viral infections, particularly of influenza virus infections. They may be administered as known in the art, e.g. intravenously, subcutaneously, intramuscularly or, most preferably, intranasally. The virus strains disclosed herein and the vaccines made thereof may, however, also be used as vectors or shuttles to present heterologous antigens to the immune system, e.g. antigens of viral envelope proteins such HIV-1 or hepatitis antigens.

Further preferred embodiments are defined in the dependent claims.

In order that the invention described herein may be more fully understood, the following Examples are set forth. They are for illustrative purposes only and are not to be construed as limiting this invention in any respect.

Example 1 : Virus production

Cultivation of Vero /SF ( = serum-free) cells:

SF-Medium: DMEM (Biochrom F0435), Ham's F1 2 (Biochrom F081 5), 5mM L-

Gln, 0.1 % SF-supplement (a) or (b); antibiotics (only for first passage of virus isolation). SF-Supplement: protein hydrolysate of non-animal origin, without functional proteins such as insulin, transferrin or growth factors: a) 62.5 g hy-soy/UF, Quest 5X59100, to 500 g HQ-water, filtered with PES 0.2 μm filter; b) 1 2.5 g hy-pep 1 51 0, Quest, to 100 g HQ-water, filtered with

PES 0.2 μm filter.

The content of a deep frozen (liquid nitrogen) disinfected (70% ethanol) ampule of WCB Vero cells was thawed and added to 9 ml of cold serum-free (SF) medium in a 10 ml tube and centrifuged for 10 min at l OOOrpm (1 70 g). The pellet was resuspended in SF-medium to a total of 30 ml, transferred to a 80 cm² Roux bottle and incubated at 37 ° C and 7%Cθ2 for at least 1 5 min. Thereafter, the medium was removed and the cells were washed with approx. 0.1 ml/cm² PBS def.( = PBS without Ca^{2 +} and Mg^{2 +} ). Addition of trypsin/EDTA-solution (8-10 μl/cm²; 0.1 % trypsin / 0.02% EDTA-solution) and incubation at room temperature for about 3 min. Detaching by gently pushing the Roux bottle against palm of the hand, addition of SF-medium and trypsin inhibitor (Sigma, T6522) at a quantity of about 1 /5 of volume of the trypsin/EDTA solution. Repartition of the cell suspension to Roux bottles or roller bottles, incubation at 37° C and 9% Cθ2-

MDCK cells were grown in DMEM/Ham's F1 2 + 2% FCS (heat inactivated); embryonated hen eggs were 1 1 -1 2 days old and of SPF (specific pathogen free) origin.

Propagation of virus strains:

Old medium from roller bottles containing Vero cells was removed and cells were infected with virus by addition of 5 ml virus suspension in SF-medium to each roller bottle, resulting in an MOl (multiplicity of infection) of approximately 0.01 . After incubation for 45 minutes at 33 °C the virus inoculum was removed with a pipette. 90ml of SF-medium supplemented with 0.5 - 10, preferably 2 - 5 and most preferably 2 μg/ml porcine trypsin (supplier: AvP) or human recombinant trypsin or trypsinogen (own production) and 0.5 g/l sodium bicarbonate were added to each roller bottle and the bottles incubated at 33 ° C and 5 % Cθ2- For the production of attenuated live vaccine samples for use in animal testing and in human clinical trials the SF-medium was supplemented with trypsin and, additionally, with Benzonase ' ^ at a concentration of 2 - 30, preferably 5 - 1 5, and most preferably 10 Units of Benzonase™ per ml of medium. Virus was harvested after 64 hours post infection by centrifugation of the culture supernatant for 5 min at 4000 rpm (3000g) at 10° C in 50 ml-tubes. The supernatant was pooled for each virus strain and stored at + 4° C. Aliquots thereof were used for vaccine testing.

For storage purposes the virus preparations may be freeze-dried and stabilizer such as, for example, trehalose and lactalbumin enzymatic hydrolysate in HEPES buffer may be added. Reconstitution may be done with sterile water. Example 2: Comparison of trypsin inactivation in cell cultures

Table 1 : Tr sin inactivation in Vero vs. MDCK cell culture

Supernatants obtained from uninfected Vero cell cultures (grown in SF medium as described in Example 1 ) and MDCK cell cultures (grown in FCS-supplemented medium as described in Example 1 ) were tested for their capacity to inactivate trypsin of different origin that has been added to the supernatant at time = 0 h at equal concentrations each. Porcine trypsin has been applied in two different qualities (obtained from different manufacturers), i.e. with high or low activity. The results are presented in Table 1 and in Figures 1 and 2.

The data unambiguously show that bovine trypsin is rapdily inactivated in Vero cell culture supernatant and less rapidly in MDCK cell culture supernatant. Porcine and human rec trypsin (manufactured in our laboratories) remain fully active in MDCK supernatants while they are gradually inactivated in Vero supernatants at approximately half or less of the velocity of bovine trypsin inactivation. The difference of the porcine trypsins tested is only in the starting OD-level at 247 nm, while the inactivation characteristics are essentially identical for both lots of porcine trypsin.

Example 3: Comparison of various viral properties after growth on different host cell substrates

Virus propagation was carried out as described in Example 1 for the different host cell substrates. Each of the seven isolates recovered on Vero cells was reactive with human erythrocytes but not with chicken erythrocytes and none of them accumulated in embryonated eggs. On the other hand, all isolates recovered on MDCK cells were reactive both with chicken and human erythrocytes and were capable of growing in eggs. Although these differences were not seen in influenza A viruses of the H 1 N 1 substype nor in influenza B isolates (see subsequent Tables 3 and 4), it may nevertheless be assumed that cultivation of influenza viruses on Vero cells will maintain antigenic properties more properly than cultivation on other substrates.

Table 2: Characteristics of H3N2 viruses isolated from clinical material on Vero/SF cells

From the data in Table 3 it appears that H 1 N 1 influenza viruses may be less susceptible to adaptive selection, as the Vero and MDCK-grown isolates do not exhibit significant differences in their hemagglutination characteristics nor in their HA sequences. A similar conclusion may be drawn for the B isolates listed in Table 4.

The clinical starting material (e.g. serum samples, swabs) for virus isolation and replication was primarily obtained from: 1 . Institute of Virology, Vienna, Austria (Prof. F. Heinz) 1995/96, 1 996/97

2. Unite de Genetique Moleculaire des Virus Respiratoires, Institute Pasteur, Paris, France (Prof. S. van der Werf) 1 996/97

3. Public Health Laboratory Service, London, UK (Dr. M. Zambon) 1 996/97

4. Laboratoire Central de Virologie, Hδpitaux Universitaires de Geneve, Geneva, Switzerland (Dr. W. Wunderli) 1 996/97, 1 997/98 5. Virus Unit, Queen Mary Hospital, Hong Kong (Dr. W.L. Lim) 1997/98

Table 3: Characteristics of H 1 N 1 viruses isolated from clinical material on Vero/SF cells

Tabelle 4: Characteristics of B viruses isolated from clinical material on Vero/SF cells

Table 5: Amino acid changes in HA, NA and M proteins of H3N2 influenza viruses isolated on different host systems

The results show that with some isolates there was no alteration of the HA sequence of Vero or MDCK propagated viruses over the HA sequence directly obtained from the swab material by PCR amplification. In some other isolates grown on MDCK cells the HA and/or NA sequences were deviating from the corresponding sequences obtained on Vero cells. The Vero-derived viruses did not show, however, any deviations in the HA sequence over the HA sequence of the swab isolates, where determined.

Table 6: Immunogenicity of Vero-, MDCK- and Egg-derived viruses for macaques

The Macaques were immunized i.n. in the absence of anesthesia with 1 ml of virus suspension V - Vero- isolated virus

M - MDCK -isolated viruses

E - egg isolated viruses

Table 7: Virulence of Vero- and MDCK- derived variants of A/Vienna/47/96 wt virus for ferrets

Animals were immunized i.n. under ether narcosis with 1 ml of virus suspension.

N- normal temperature from 38.1 ° C to 39.9 ° C;

F- fever, more than 40.0° C. The most surprising, yet important result in Table 6 is the very low immunogenicity of MDCK-derived A/Vienna/47/96 virus compared with the corresponding Vero-derived virus. It is no particular surprise that the egg-derived viruses show only poor immunogenicity.

Similarly, the results listed in Table 7 indicate that Vero-derived viruses are less, if at all, altered by adaptive selection on their host substrate in comparison to MDCK-derived viruses. This means that relative to the MDCK-derived viruses the Vero-derived viruses maintain more or even all of the immunologically relevant, particularly antigenic, properties of the original virus.

Example 4: Vaccine production with preferred strains

The process described in Example 1 was also used for the production of vaccine samples for animal testing and human clinical studies. It is understood that the process of virus propagation described therein also encompasses variations that could be suggested or applied by a person of ordinary skills in the art without inventive input and as long as the variations do not change the sense of the present invention as described herein and in the claims.

Vaccine samples containing one or more of the preferred influenza A or B wildtype strains, master strains or reassortant strains (that are subsequently described in more detail) were exclusively produced using the continuous Vero cell line as the host cell system (unless for purposes of comparison with samples obtained from other host substrates) in serum-free medium additionally supplemented with the nutritional ingredients and enzymes as described in Example 1 .

Some methods suitable for modifying wildtype viruses including the methods of attenuation (e.g., temperature sensitivity), cold adaptation and reassortment are known in the art and extensively reviewed, for instance, in WO 99/64068.

Further characteristics of the two most preferred influenza A and B master strain candidates useful for attenuated live vaccine production, e.g., by 6/2 reassortment with the HA and NA genes of actual epidemic influenza viruses recommended by the WHO, are given in the following Tables 8 - 13. Table 8: Characteristics of master strain candidates for live influenza vaccines

Table 9: Full Sequence of the 8 genome segments and of the 10 corres onding proteins of strain A/Sin a ore/1 /57/ca

ca - cold adapted

It shall be noted, however, that the genome segments No. 4 and 6, i.e., the HA and NA genes, are not required to characterize the influenza A master strain candidates, because these genes will be exchanged for the corresponding genes of actual epidemic influenza viruses (as mentioned hereinbefore). The features important for the safety of a vaccine, e.g. temperature sensitivity, or features that allow intranasal administration of a vaccine, namely cold adaptation (because the average temperature in a nose is lower than the usual body temperature), are primarily caused by mutations in the remaining 6 genome segments.

The following Table 10 lists the mutations in the genome segments of A/Singapore/1 /57/ca compared to the corresponding wildtype strain A/Singapore/1 /57/wt.

Table 10: Mutations in the genome segments of attenuated, temperature sensitive, cold adapted influenza strain A/Singapore/1 /57/ca compared to A/Singapore/1 /57/wt strain

Total number of mutations - 1 3 (8 coding) * coding mutations Preferred variants of A/Sing/1 /57/ca comprise the ones listed in the following Table 1 1 , wherein "Δ" means "del" or "delta" and stands for a mutant that contains at least one "deletion" in its NS gene segment.

* genome segment originating from A/Singapore/1 /57/ca * * genome segment originating from influnza A/PR8/34 ca - cold adapted; ts - temperature sensitive; aa - amino acid(s) IFN-induct. - strain causes interferon release in host substrates that are able of

IFN production, as well as in animal or human immune systems upon administration. IFN-sensit. - strain is sensitive towards interferon; replication in IFN producing systems is reduced or stopped. Sing ca/ΔNS 87 - strain A/Singapore/1 /57/ca containing deletion of 87 amino acids in NS1 gene at aa position 36-1 23. Sing ca/ΔNSPR8 - strain A/Singapore/1 /57/ca containing the NS gene segment from A/PR8/34 (herein also abbreviated "PR8" ) which contains a deletion of the entire NS1 gene. Sing ca/NS1 24PR8 - strain A/Singapore/1 /57/ca containing the NS gene segment from A/PR8/34 which contains a stop codon at aa position 1 24 of the NS1 gene. + /- means that the phenotype needs further clarification and can not yet be unambiguously defined.

The following Tables 1 2, 1 3 and 1 3A refer to preferred influenza B master strain candidates and to variations and reassortants, respectively, thereof.

Table 12: Full Sequence of the 8 genome segments and of the 1 1 corresponding proteins of strain B/Vienna/1 /99/ca

ca - cold adapted

The original strain B/Vienna/1 /99 was isolated on Vero cell culture grown with serum-free medium in February 1 999 in Vienna, Austria from a 1 2 year old female with acute influenza. It was rated as B/Beijing/184/93-like by the Center for Disease Control (CDC), Atlanta, USA. After an additional passage at 33° C the wildtype strain - designated as B/Vienna/1 /99 wt - was attenuated by 22 serial passages at 25 °C using the same cell culture system. The plaque purification was done at 25 ° C for the first and at 33 ° C for the following four rounds. The derived plaque purified clone was amplified and stored at -70°C, designated as B/Vienna/1 /99 ca or briefly BV22. The identity as a B/Beijing/184/93-like virus was confirmed by Hl-assay with standard anti-serum from NIBSC.

In addition, further passaging of strain B/Vienna/1 /99 ca for 1 5 additional passages (i.e. a total of 37 passages on serum-free Vero cell culture) resulted in a mutant B/Vienna/1 /99 ca37 (abbreviated BV37) with properties even superior to the ones of BV22. This mutant contains an increased number of mutations vis-a-vis BV22 and appears to be the currently most promising candidate for the production of a whole-virus vaccine, particularly for an attenuated influenza live vaccine, based on a non-recombinant influenza virus mutant. The additional mutations are listed in Table 1 3A below:

Table 1 3 A: Mutations for BV22 and BV37 compared to B/Vienna/1 /99 wt 1 st passage on Vero/SF

Comparison with influenza sequence database 1 3.2. 2001 (www.flu-lanl.gov): a) unique mutations underlined in bold type; b) mutations common with:

^* B/Lee/40, B/Osaka/70, B/Kadoma/1 076/99 (resulting amino acid: I) + B/Lee/40, B/Osaka/70 ^# often: B/Lee/40, B/Ann Arbor/1 /66 ca & wt, B/Singapore/222/79, B/North Dakota/83, B/Norway/1 /84, B/Ibaraki/2/85, B/Ann Arbor/1 /86, B/Victoria/2/87, B/Aichi/5/88

• B/Kanagawa/73 It shall be understood that the influenza A and B master strains according to the present invention shall not be limited to the features and genetic characteristics explicitly listed in the tables herein but shall also comprise minor variations thereof as long as such variations are in the sense of the present invention and do not subtantially alter any one of the functional features of the virus. Such variations may occur, for instance, due to additional steps of virus multiplication or propagation (e.g. for the purpose of obtaining material for sequence analyses).

Moreover, the gene sequences listed herein include the primer sequences (located at the beginning and at the end of each genome segment) that were used along with the present invention, which primer sequences may differ from the corresponding true sequences of the viral genome segments of either or both the wildtype and the attenuated virus strains.

Example 5: Vaccine safety and efficacy

The subsequent data confirm temperature sensitivity and vaccine safety for influenza vaccines manufactured according to the present invention, e.g., as described in Example 1 .

Table 1 4: Antibody response of mice after one intranasal immunisation without narcosis

1 - number of animals with positive HI titer > 1 :4

2 - number of animals without detectable virus in the lungs 3- Geometric mean titer of antibodies in serum

PR8wt - influenza strain A/PR/8/34 wildtype (H 1 N1 ), pathogenic for mice PR8/Sing ca-2/6 - is the reassortant between attenuated influenza strain

A/Sing/1 /57 ca and PR8 wt, containing 2 genes (HA and NA) from PRδwt virus and all other genes from A/Sing/1 /57 ca.

PR8/Sing-ΔNS contains HA and NA genes from PR8wt, five genes from A/Sing/1 /57 ca and the NS gene of PR8 origin lacking the NS1 coding sequence (NS1 deletion or knockout).

Table 1 5: Antibody response and protection of mice after intranasal immunisation with different variants of A/Singapore/1 /57 virus (under narcosis)

¹ - number of animals with positive HI titer > 1 :4

² - va- Vero-adapted

³ - ca - cold-adapted - number of animals without detectable virus in the lungs

Table 1 6: Reproduction of wt, va and ca variants of A/Singapore/1 /57 in mouse lungs^a

Viruses Virus titer in mouse lungs post infection on day,

PFU/ml^b

2 4 6

A/Singapore/1 /57/wt 1 .6x1 0⁶ 2.2x10⁵ 1 .4x10³

A/Singapore/1 /57/wt va 2.5x1 0⁶ 2.1 x10⁶ 1 .0x10²

A/Singapore/1 /57/ca < 10 < 10 < 10 ^a Mice were infected i.n. with 50 μl of virus fluid with a titer 1 .0 x 10⁶ PFU/ml. ^b PFU/ml of 10% tissue suspension, titrated on MDCK cells. Table 1 7: Virulence of wt and ca variants of A/Singapore/1 /57 virus for ferrets

Rectal temperature of animals was recorded twice a day and characterized as follows:

N - normal temperature from 38. 1 ° C to 39.9 °C

F - fever, more than 40.0° C.

Each group consisted of 3 animals, which were immunized i.n. under ether narcosis with 1 ml of virus fluid with a titer of 2x10⁶ PFU/ml.

Table 1 8: Reproduction of 2/6 reassortant of A/Hong Kong/1035/98 wt and A/Sin a ore/1 /57/ca in mouse lungs³

^a Mice were infected i.n. under ether narcosis with 50 μ\ of virus fluid. ^b PFU/ml of 1 0% tissue suspension, titrated on Vero/SF cells, data are given as mean value for 6 mice (the lungs of each animal were treated separately). The reassortant contains the HA and NA genes from A/Hong Kong/1035/98 wt wildtype and the other 6 genes from A/Singapore/1 /57/ca.

Table 1 9: Virulence of 6/2 reassortant of A/Vienna/47/96 wt and A/Singapore/1 /57/ ca for ferrets

Viruses Virus Number of animals with fever on day subtype 1 Rhinitis"

Master strain A/Singapore/1 /57/ ca H2N2 NNN NNN NNN +

Epidemic virus A/Vienna/47/96 wt H3N2 NNN FFF FFF + + +

Reassortant A/Singapore/1 /57/ca x H3N2 NNN NNN NNN + Vienna/47/ 96 wt

Animals were immunized i.n. under ether narcosis with 1 ml of virus, 2x10⁶ PFU/ml. N- normal temperature from 38.1 ° C to 39.9 ° C; F- fever, more than 40.0°C. ^b + + + - severe rhinitis ± absence of rhinitis

The results presented in Tables 1 6 to 19 clearly demonstrate the safety of the vaccines containing the attenuated, temperature sensitive master strain or, in case of reassortants, of the vaccines based on the reassorted viruses composed of the "backbone" of the attenuated, temperature sensitive master strain (6 genes) and the HA and NA genes from, e.g., the pathogenic wildtype strain A/Hong Kong/1035/98 wt.

Table 21 : Genetic stability of the ts phenotype of B/Vienna/1 /99 ca

The strain BV22 was passaged five times at high MOl on Vero cells. Then the ts-phenotype was controlled again. The strain remained tmperature senssitive as can be seen in Table 21 .

* 9 OF1 mice per strain were immunized intranasally under ether narcosis with 10⁵ PFU. At the indicated days post infection 3 mice per group were sacrificied. Lungs and nasal turbinates were homogenized for a 1 0% (w/v) suspension in PBS def. A plaque assay of the suspensions was performed.

The data show that moderate reproduction of the ca master strain candidate BV22 was possible in the nasal mucosa while the ts property of the virus prevented reproduction in the lungs.

Table 23: Ts and ca phenotype of the reassortant influenza B strain

A 6/2 reassortant strain containing HA and NA of the wild type influenza strain B/USSR/69 wt and the other 6 genome segments from B/Vienna/1 /99 ca (BV22) was established. The origin of the hemagglutinin was tested by Hl- assay, all other genome segments by RT-PCT and restriction analysis using methods known in the art.

Table 24: Virulence of the reassortant influenza B strain in mouse lun s

* 9 OF1 mice per strain were immunized intranasally under ether narcosis with 10⁵ PFU. At the indicated days post infection 3 mice per group were sacrificied. Lungs and nasal turbinates were homogenized for a 10% (w/v) suspension in PBS def. A plaque assay of the suspensions was performed.

Example 6: Clinical study

The following vaccines (in the form of nasal sprays) were produced according to the present invention (e.g. as described in Example 1 ) for intranasal delivery. Composition per ml (after reconstitution of freeze-dried material): (1 ) Placebo: 2x SF-medium, 40mM HEPES buffer, 8% lactalbumin enzymatic hydrolγsate, 4% trehalose; (2) VeroVac H 1 : A/Beijing/262/95 (H 1 N 1 )-like preparation comprising

4.3x1 0⁷ TCID₅₀ of 6/2 reassortant A/Singapore/1 /57/ca with A/Hong

Kong/1035/98; 2x culture supernatant, 40mM HEPES buffer, 8% lactalbumin enzymatic hydrolysate, 4% trehalose; (3) Vero Vac H3: A/Sidney/5/97 (H3N2)-like preparation comprising 2.1 x10⁷ TCID₅₀ of 6/2 reassortant A/Singapore/1 /57/ca with A/SW/7729/98; 2x culture supernatant, 40mM HEPES buffer, 8% lactalbumin enzymatic hydrolysate, 4% trehalose; (4) Russian trivalent vaccine (live influenza vaccine for adults):

A/1 7/Beijing/95/25 (H 1 N 1 ) 1 .1 x10⁸ EID₅₀ A/1 7/Sidney/97/76 (H3N2) 2.3x10⁷ EID₅₀

B/60/Petersburg/95/20 1 .1 x10⁷ EID₅₀

(5) Monovalent Vero vaccine BV22: B/Beijing/1 84/93 - like preparation comprising 2x10⁶ TCID₅₀ of master strain candidate B/Vienna/1 /99/ca ( = BV22); 2x culture supernatant, 40mM HEPES buffer, 8% lactalbumin enzymatic hydrolysate, 4% trehalose;

The vaccines were administrated to 1 3 volunteers per each vaccination group. 550 μl of reconstituted vaccine (or placebo, respectively) were given intranasally to each patient on day 0 and for a second time on day 22 ± 1 . The results are summarized in Table 25 below.

Safety results:

The total number of adverse events (AE) during five days after the first and second vaccination was 14 including 9 mild and 4 moderate AE. Only one volunteer showed severe AE, comprising an increase in body temperature up to

38.8° C within 3 hours after the first vaccination without any local or systemic symptoms. During the next four hours his temperature became normal again.

After the first vaccination ^"7 AE were observed. One of them was local and six were systemic. After the second vaccination 2 local and 5 systemic AE were observed.

No significant difference in terms of safety was revealed between the groups of the study including the one with placebo. No serious AE related to the vaccination were observed except for the one mentioned above. Two of the moderate AE occurred in the H3N2 group (temperature elevation up to 37.6° and acute pharyngitis on day 3 in one volunteer; nasal obstruction, discomfort in the throat on day 22-24 and temperature elevation up to 37.5 ° C in another volunteer), and one in the H 1 N 1 group (pain in the throat, rhinitis from day 22- 26, temperature elevation up to 37 - 37.8 ° C between days 22-24).

(8) patient developed spontaneous infection during course of study.

The results obtained from the clinical study thus confirm a very good safety of the vaccines produced according to the present invention and using the preferred influenza A and B master strain candidates of the present invention.

Claims

CLAIMS We claim

1 . A method for the manufacture of a whole-virus vaccine, preferably an attenuated live vaccine, comprising the steps of: a) infecting African Green Monkey Kidney (Vero) cells with a desired virus, wherein the Vero cells have been grown in and separated from a serum-free medium that is also free of non-serum proteins; b) combining the infected cells with a suitable serum-free cell culture medium that is also free of non-serum proteins except for a protease and a nuclease; and c) incubating the cells in the presence of said protease and said nuclease to allow for production of infectious virus and, simultaneously, for digestion of nucleic acid material released to the cell culture medium; d) harvesting infectious virus by collecting virus-containing supernatant obtained from centrifugation of the cell culture; and e) preparing a vaccine thereof comprising subjecting the virus-containing supernatant to at least one processing step selected from the group consisting of filtering, concentrating, freezing, freeze-drying, and stabilizing by addition of a stabilizing agent.

2. The method according to claim 1 , which does not involve a step of protein separation or purification.

3. The method according to claim 1 or 2, which does not involve a step of chromatographic separation or purification, and preferably does not contain any purification step other than centriguation and/or filtration.

4. The method according to any one of claims 1 to 3, which comprises at least one step of sterile filtration of the virus-containing supernatant.

5. The method according to any one of claims 1 to 4, wherein the nuclease has DNAse and/or RNAse activity, and preferably is Benzonase.

6. The method according to any one of claims 1 to 5, wherein the protease and the nuclease are added to the cell culture medium once prior to or at the beginning of incubation of the infected cells.

7. The method according to any one of claims 1 to 6, wherein the protease comprises trypsin and/or trypsinogen of human recombinant or porcine origin which is present in the cell culture medium at an initial concentration of 0.5 - 10, preferably 2 - 5 μg per ml medium.

8. The method according to any one of claims 1 to 7, wherein the cell culture medium comprises nuclease at an initial concentration of 2 to 30, preferably 5 to 1 5, U per ml of medium.

9. The method according to any one of claims 1 to 8, wherein the incubation in step (a) is carried out for 10 to 1 20 minutes, preferably for 30 to 60 minutes.

10. The method according to any one of claims 1 to 9, wherein the virus is selected from the group consisting of a wildtype virus, a primary isolate directly obtained from an infected individual, a recombinant virus, an attenuated virus, a Vero adapted virus, a cold-adapted virus, a temperature-sensitive virus, and a reassortant virus.

1 1 . The method according to any one of claims 1 to 10, wherein the virus is an influenza A virus, preferably of subtype H3N2 or H 1 1 , or an influenza B virus.

12. The method according to any one of claims 1 to 1 1 , wherein the virus has an interferon inducing and/or interferon sensitive phenotype.

1 3. The method according to any one of claims 1 to 1 2, wherein the virus is an influenza virus selected from the group consisting of strains A/Sing/1 /57ca, A/Sing/1 /57ca/ΔNS 87, A/Sing/1 /57ca/ΔNSPR8, A/Sing/1 /57ca/NS1 24PR8, B/Vienna/1 /99ca, B/Vienna/99/ca37 and any attenuated variants and reassortants derived from any one of these strains.

14. A whole-virus vaccine, preferably an attenuated live vaccine, characterized in that in its ready-for-use form it comprises essentially unmodified, optionally filtered and/or concentrated, virus-containing supernatant of a serum-free and protein-free Vero cell culture used for production of said virus.

1 5. The vaccine according to claim 1 4, characterized in that it selectively agglutinates human erythrocytes but not chicken erythrocytes.

1 6. The vaccine according to claim 14 or 1 5, characterized in that it contains a suitable stabilizing agent.

1 7. The vaccine according to any one of claims 14 to 1 6, characterized in that it is in the form of a liquid, freezed or freeze-dried preparation, optionally suitable for intranasal delivery.

1 8. The vaccine according to any one of claims 14 to 1 7, characterized in that it is a live attenuated vaccine, preferably comprising whole influenza virus.

1 9. The vaccine according to any one of claims 14 to 1 8, characterized in that it comprises at least one influenza virus having a phenotype with one or more characteristics selected from the group consisting of cold adapted, temperature sensitive, interferon inducing, interferon sensitive.

20. The vaccine according to claim 18, wherein the influenza virus is selected from the group consisting of strains A/Sing/1 /57ca,

A/Sing/1 /57ca/ΔNS 87, A/Sing/1 /57ca/ΔNSPR8, A/Sing/1 /57ca/NS1 24PR8, B/Vienna/1 /99ca, and any attenuated variants and reassortants derived from any one of these strains.

21 . The vaccine according to claim 14, obtainable by a method of manufacture as defined in any one of claims 1 to 13.

22. A whole-virus vaccine, preferably an attenuated live vaccine, comprising at least one influenza virus selected from the group consisting of strains A/Sing/1 /57ca, A/Sing/1 /57ca/ΔNS 87, A/Sing/1 /57ca/ΔNSPR8,

A/Sing/1 /57ca/NS1 24PR8, B/Vienna/1 /99ca, and any attenuated variants and reassortants derived from any one of these strains.

23. The vaccine according to claim 21 , characterized in that it selectively agglutinates human erythrocytes but not chicken erythrocytes.

24. The vaccine according to claim 22 or 23, obtainable by a method of manufacture according to any one of claims 1 to 13.

25. Use of a vaccine defined in any one of claims 1 4 to 24 for prophylactic or therapeutic administration against viral infection.

26. Use of at least one influenza virus selected from the group consisting of strains A/Sing/1 /57ca, A/Sing/1 /57ca/ΔNS 87, A/Sing/1 /57ca/ΔNSPR8, A/Sing/1 /57ca/NS1 24PR8, B/Vienna/1 /99ca, and any attenuated variants and reassortants derived from any one of these strains, for the manufacture of a vaccine, preferably for the manufacture of a live attenuated influenza vaccine.